Hierarchy helps recognize ancient shorthand symbols in manuscripts
Evaluating Hierarchy-Aware Deep Learning for the Recognition of Tironian Notes
Computer Vision and Pattern Recognition
Summary
Reading ancient shorthand symbols called Tironian notes is very hard because there are many similar-looking signs. The authors studied whether understanding the relationships between these signs can help computers recognize them better. They compared different deep learning models that either treat the signs as separate classes or use a hierarchy of sign groups. Their tests show that models using the hierarchy perform better when there's no extra tuning to a specific manuscript, while simpler models work better after some adaptation. This means that knowing how these shorthand signs relate can help machines read them, especially in challenging cases.
What this means in practice
- •For digital archivists: Automatically identify Tironian shorthand signs in digitized Latin manuscripts to speed up text analysis without extensive manual labeling.
- •For document analysis engineers: Integrate hierarchy-aware deep learning models to improve recognition of complex symbol sets where training data is limited or varied.
Authors
Yule Kang, Thomas Gorges, Janne van der Loop, Franziska Marske, Nikolaus Weichselbaumer, Tino Licht, Vincent Christlein
Abstract
Tironian notes are generally regarded as the first Latin shorthand system and are notable for their large, fine-grained symbol inventory. Their high visual similarity and large class set make manual reading time-consuming, leaving manuscripts that contain Tironian notes inaccessible to many researchers. Automatic recognition is also challenging because models must distinguish subtle differences in stroke shape and sign structure while realistic training data remain scarce. However, standard flat classifiers do not explicitly use visual or structural relations between related signs. This paper investigates whether structural relationships between Tironian notes can support automatic recognition. We use the Supertextus Notarum Tironianarum (SNT) by Martin Hellmann, which provides idealized sign forms and a hierarchical organization of Tironian notes. We compare flat ResNet18, ConvNeXt, Shifted Window Transformer (Swin), and Vision Transformer (ViT) classifiers with Hierarchical Deep Convolutional Neural Network (HD-CNN)-style coarse-to-fine models and hierarchy-aware routing models based on visual class cleaning and similarity-based re-clustering. The models are evaluated on handwritten samples and manuscript-domain samples from Vergilius Turonensis, both with and without limited few-shot adaptation to the manuscript domain. The results show that the relative performance of flat and hierarchical models depends on adaptation. On Vergilius Turonensis, HD-CNN achieves the best non-adapted Top-1 result with 45.43%, while flat classification reaches the best Top-1 result after few-shot adaptation with 82.09%. Overall, the results indicate that hierarchical structure can support Tironian note recognition, especially under non-adapted conditions.